The dimeric versus monomeric status of 14-3-3zeta is controlled by phosphorylation of Ser58 at the dimer interface.

The 14-3-3 proteins play a central role in the regulation of cell growth, cycling, and apoptosis by modulating the functional activities of key signaling proteins. Through binding to a phosphoserine motif, 14-3-3 alters target proteins activities by sequestering them, relocalizing them, conformationally altering their functional activity, or by promoting interaction with other proteins. These functions of 14-3-3 are facilitated by, if not dependent on, its dimeric structure. We now show that the dimeric status of 14-3-3 is regulated by site-specific serine phosphorylation. We found that a sphingosine-dependent kinase phosphorylates 14-3-3 in vitro and in vivo on a serine residue (Ser58) located within the dimer interface. Furthermore, by developing an antibody that specifically recognizes 14-3-3zeta phosphorylated on Ser58 and employing native-PAGE and cross-linking techniques, we found that 14-3-3 phosphorylated on Ser58 is monomeric both in vitro and in vivo. Phosphorylated 14-3-3 was detected solely as a monomer, indicating that phosphorylation of a single monomer within a dimer is sufficient to disrupt the dimeric structure. Significantly, phosphorylation-induced monomerization did not prevent 14-3-3 binding to a phosphopeptide target. We propose that this regulated monomerization of 14-3-3 controls its ability to modulate the activity of target proteins and thus may have significant implications for 14-3-3 function and the regulation of many cellular processes.

The 14-3-3 proteins belong to a highly conserved family of phosphoserine binding proteins that regulate multiple signaling pathways involved in the control of cell division, growth, and apoptosis (1-4). 14-3-3 proteins influence the function of bound phosphoserine proteins in a variety of ways: sequestering them from cellular targets, controlling their enzymatic activities, relocating them, or acting as adapter molecules in mediating the association of two distinct client proteins (2,4). To date, 14-3-3 proteins have been found to bind to over 100 cellular "client" proteins and have been demonstrated to participate in multiple signaling pathways and biological functions (1)(2)(3)(4)(5). Seven isoforms of 14-3-3 have been described in humans, ␤, ⑀, ␥, , , , and , which are expressed from separate genes and are relatively abundant in all tissues (with the exception of and that are predominantly expressed in epithelium and T-cells, respectively). In most cases, the regulated phosphoryl-ation of the client protein controls their binding to 14-3-3 proteins. However, the regulation of the 14-3-3 proteins themselves by phosphorylation and the consequences are virtually unknown.
14-3-3 proteins are dimeric, being composed of two 30-kDa monomers that are each capable of binding a phosphoserine motif. Each monomer is composed of nine ␣ helices arranged in antiparallel fashion to form an amphipathic groove that mediates phosphoserine target binding. The resolved crystal structures of 14-3-3 and 14-3-3 demonstrate the structural similarity between isoforms and explain the high degree of conservation across the family, with conserved residues lining the phosphoserine binding groove (6,7). The N-terminal helices of 14-3-3 mediate dimer formation with helix 1 of monomer A interacting with helices 3 and 4 of monomer B. Hetero-as well as homodimerization of isoforms occurs (8,9). However, it is not known whether 14-3-3 monomers exist in vivo and if so what control mechanisms may exist to regulate their formation.
It has remained a mystery exactly how 14-3-3 can regulate so many client proteins with such diverse functions. A recently proposed model provides a possible explanation (5), describing 14-3-3 as a molecular anvil that deforms a bound phosphoserine client protein, thus affecting its conformation, altering enzymatic activity, or masking/revealing phosphorylation sites or nuclear transport sequences. This deforming activity of 14-3-3 is wholly dependent on 14-3-3 protein being dimeric and binding to multiple sites in the client protein, although initially a single 14-3-3 binding motif in the client protein may act as a dominant or "gatekeeper" site for 14-3-3 interaction. In the context of this model, monomeric 14-3-3 would retain binding to a gatekeeper site but would not exhibit any "molecular anvil" activity and would therefore be unable to perform deforming functions.
Previous studies demonstrated that a sphingosine-dependent kinase (SDK) 1 isolated from Balb/c 3T3 fibroblasts phosphorylated 14-3-3 proteins in vitro on Ser 58 in helix 3; however, the effect of this phosphorylation was not known (10). From the crystal structure of 14-3-3 this residue is located in the dimer interface just 5.8 Å from a conserved arginine (Arg 18 ) in helix 1 of the opposing monomer. We have now determined the functional consequence of Ser 58 phosphorylation on the structure of 14-3-3 and show that phosphorylation of this residue disrupts dimer formation. Moreover, 14-3-3 phosphorylated on Ser 58 is detected in vivo in response to sphingolipid and is also found to be a monomer. These results demonstrate for the first time that the dimeric status of 14-3-3 can be regulated in vivo by sitespecific phosphorylation at the dimer interface, a finding that * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ To whom correspondence should be addressed. may have profound implications for 14-3-3 function and the regulation of many cellular processes.

EXPERIMENTAL PROCEDURES
Sphingolipids and Antibodies-Sphingolipids were purchased from Biomol. Anti-14-3-3 antibody was raised in New Zealand White rabbits using glutathione S-transferase-14-3-3 (GST-14-3-3) as the immunogen. Polyclonal anti-14-3-3 antibodies were subsequently purified from rabbit serum using a GST-14-3-3 affinity column and used for immunoblotting at a dilution of 1:2000. Anti-phospho-Ser 58 14-3-3 antibody (anti-pS58) was raised by immunizing New Zealand White rabbits with the phosphorylated CGARRSpSWRVVS peptide conjugated to keyhole limpet hemocyanin. The phospho-specific Ser 58 antibodies were subsequently affinity purified from rabbit serum using the phospho-and an unphosphorylated form of the immunizing peptide, essentially as described previously (11). The purified anti-phospho-Ser 58 was used for immunoblotting at a concentration of 1 g/ml. Horseradish peroxidase-conjugated anti-rabbit antibody was purchased from Pierce.
In Vitro SDK Reactions-An S100 cytosolic fraction was prepared from serum-starved, PMA-stimulated Balb/c 3T3 fibroblasts as described previously (10). Aliquots of extract were frozen away at Ϫ70°C and were subsequently used in SDK reactions to phosphorylate 14-3-3. SDK assays were carried out essentially as described previously (10) with the following changes: assays were carried out in 20-l volume with 100 ng of recombinant 14-3-3. S100 extract made up half the assay volume, and SDK assay buffer was 20 mM Tris-Cl, pH 7.4, 15 mM MgCl 2 , 25 M ATP, 3 mM dithiothreitol with or without 2 Ci of [␥-32 P]ATP. Sphingolipid was added to a final concentration of 100 M in 0.5% w/v n-octyl glucoside. Reactions were incubated at 37°C for 15 min.
Cross-linking Studies-14-3-3 phosphorylated by SDK was subjected to cross-linking with bis(sulfosuccinimidyl) suberate (BS 3 ) (Pierce). BS 3 (50 g/ml) was added to an SDK reaction after incubation and the mixture incubated for a further 5 min at 37°C. Cross-linking was terminated by addition of ethanolamine to a final concentration on 100 M. Cross-linked samples were subjected to SDS-PAGE, Western transferred, and immunoblotted as indicated.
Cell Stimulation and Lysate Preparation-Balb/c 3T3 cells were stimulated at 80% confluence in 10-cm dishes. After stimulation as indicated, cells were put on ice and washed twice with phosphatebuffered saline. Cells were released from the dishes by scraping and pelletted by centrifuging at 1000 ϫ g for 5 min at 4°C. The cell pellets were lysed on ice in 100 l of Nonidet P-40 lysis buffer containing protease inhibitors (10 g/ml leupeptin, 2 mM phenylmethylsulfonyl fluoride, and 10 g/ml aprotinin) and phosphatase inhibitors (2 mM sodium vanadate, 1 mM sodium fluoride, 1 mM sodium molybdate, 1 mM sodium pyrophosphate, and 1 mM ␤-glycerophosphate). Cell debris was removed by centrifugation at 10,000 ϫ g for 15 min at 4°C and protein concentration determined by BCA assay (Pierce). 50 g of lysate was subjected to native-PAGE and immunoblotted as indicated. CTLL-2 cells (2 ϫ 10 6 per treatment) were stimulated as indicated and after pelletting by centrifugation at 400 ϫ g for 5 min were washed with phosphate-buffered saline containing 1 mM ␤-glycerophosphate and lysed on ice in 50 l of Nonidet P-40 lysis buffer containing protease and phosphatase inhibitors. Cell debris was removed by centrifugation at 10,000 ϫ g for 15 min at 4°C, and 25 l of lysate was subjected to native-PAGE and immunoblotted as indicated.
PAGE and Immunoblotting-Native-PAGE was carried out on 12.5% PAGE gels with 5% stacking gels. Gels were prepared using the standard Laemmli recipe but excluding SDS. SDS was also excluded from the sample load buffer. Benchmark prestained protein markers (Invitrogen) were used as an indicator of protein migration. All gels were transferred to nitrocellulose by electroblotting using standard buffers, and blots were blocked in TNT (10 mM Tris-Cl, pH 8, 150 mM NaCl, and 0.05% (v/v) Tween 20) containing 1% (w/v) blocking reagent 1096 176 (Roche Applied Science). Immunoblotting was carried overnight at 4°C with anti-phospho-Ser 58 antibody and for 1 h at room temperature with the anti-14-3-3 antibody. All washes were performed in TNT with 0.5% (w/v) blocking reagent added to antibody incubations. Immunoblots were developed by enhanced chemiluminescence using proprietary reagents (Amersham Biosciences).

The Phosphorylation of 14-3-3 on Ser 58 by an SDK Activity Disrupts 14-3-3 Dimers-A SDK activity that phosphorylates
Ser 58 of 14-3-3 was previously identified in cytosolic extracts of PMA-stimulated Balb/c 3T3 fibroblasts (10). We have reproduced the in vitro SDK assay system described in the original report. As shown previously (10), phosphorylation of recombinant wild type 14-3-3 was induced by sphingosine (data not shown) and the sphingosine analogue, dimethyl sphingosine (DMS) (Fig. 1A) but not by sphingosine 1-phosphate (data not shown). Furthermore, recombinant 14-3-3 mutated at Ser 58 is not phosphorylated in response to DMS confirming that Ser 58 is the sole site of phosphorylation by SDK (Fig. 1A). We have now raised a polyclonal antibody to a phosphopeptide corresponding to the amino acid sequence encompassing phospho-Ser 58 14-3-3: CGARRSpSWRVVS, where pS represents phosphoserine. The antibody specifically recognized the phosphorylated, but not the unphosphorylated, form of the Ser 58 peptide nor a phosphopeptide corresponding to another known phosphorylation site in 14-3-3, Ser 185 (Fig. 1B). Using the in FIG. 1. 14-3-3 phosphorylated in vitro on Ser 58 by a sphingolipid-induced activity is specifically recognized by an anti-phosphopeptide antibody corresponding to Ser 58 of 14-3-3. A, S100 extract from PMA-stimulated Balb/c 3T3 fibroblasts was incubated either without (Ϫ) or with wild type (ϩWt) or mutant (ϩS58E) recombinant 14-3-3 (r14-3-3) in the presence or absence of DMS for 15 min at 37°C and proteins separated on 12.5% SDS-PAGE. Phosphorylated proteins were detected by [␥-32 P]ATP incorporation (upper panel), and loading of r14-3-3 substrate was determined by Coomassie staining (lower panel). B, phosphopeptides (p) and non-phosphopeptides (no prefix) encompassing Ser 58 and Ser 185 of 14-3-3 immobilized on nitrocellulose were immunoblotted with an anti-Ser 58 phosphopeptide antibody. C, as in A, wild-type r14-3-3 was incubated with S100 either in the absence (Ϫ) or presence (ϩ) of DMS separated by SDS-PAGE and immunoblotted (IB) with either anti-phospho-Ser 58 (anti-pS58) or anti-14-3-3 antibody.
14-3-3 proteins are characteristically acidic proteins with pI values of between 4 and 5. This being so, they resolve well on native-PAGE when run under neutral conditions. Native-PAGE has been employed previously in 14-3-3 studies to distinguish between monomeric and dimeric 14-3-3 forms (13). Using native-PAGE in conjunction with immunoblotting with the Ser 58 phosphopeptide-specific polyclonal antibody, we have examined the effect of Ser 58 phosphorylation on 14-3-3 dimerization. In vitro SDK reactions were carried out either in the presence or absence of added recombinant 14-3-3, subjected to native-PAGE, and immunoblotted after Western transfer with either the phospho-specific Ser 58 antibody ( Fig. 2A) or with a polyclonal anti-14-3-3 antibody (Fig. 2B). Phosphorylation of both endogenous and recombinant 14-3-3 was detected with the Ser 58 phospho-specific antibody only when DMS was present in the kinase reaction, corresponding to stimulation of the SDK activity ( Fig. 2A). Prestained markers were used as an aid to compare the migration of immunoreactive species between blots. The mobility of the phosphorylated 14-3-3 coincides with the mobility of the seventh protein marker, whereas immunoblotting of the same reactions with anti-14-3-3 indicates that both the endogenous and recombinant 14-3-3 migrate with the sixth protein marker (Fig. 2B). Furthermore, this mobility coincides with the mobility of purified recombinant 14-3-3 as demonstrated by Coomassie Blue staining of native gels (Fig.  2C). In addition, the anti-14-3-3 antibody also detected a small amount of protein running with the seventh marker that corresponds to the protein detected by the anti-phospho-specific Ser 58 antibody (Fig. 2B), identifying the faster migrating species as 14-3-3. This increase in mobility detected by the phospho-specific antibody is greater than can be accounted for by the addition of a phosphate group alone which adds two extra negative charges and 80 Da molecular mass to the protein. This implies that the difference between the two differentially migrating 14-3-3 species is due to conformational effects and strongly suggests that 14-3-3 phosphorylated on Ser 58 by SDK is monomeric, whereas the unphosphorylated 14-3-3 is dimeric. Furthermore, no phosphorylated 14-3-3 is detected migrating at the sixth marker ( Fig. 2A), suggesting that phosphorylation of a single monomer within a 14-3-3 dimer is sufficient to disrupt the dimeric structure of 14-3-3.
Demonstration That 14-3-3 Phosphorylated on Ser 58 Is Monomeric by Cross-linking-We employed cross-linking to examine the dimeric status of Ser 58 -phosphorylated 14-3-3 compared with unphosphorylated 14-3-3. 14-3-3 was phosphorylated in vitro in an SDK reaction and then subjected to cross-linking with BS 3 . Cross-linking was terminated using ethanolamine and the cross-linked complexes separated by SDS-PAGE. Gels were Western transferred and immunoblotted either with Ser 58 phospho-specific or anti-14-3-3 antibody (Fig. 3). As can be seen, the anti-14-3-3 antibody detected protein of molecular mass 28 kDa, corresponding to the size of denatured monomeric 14-3-3 and also a protein of 56 kDa when the SDK reaction was treated with BS 3 , corresponding in size to covalently cross-linked 14-3-3 dimers (Fig. 3). In contrast, the phospho-specific antibody only detected protein of the size of denatured monomeric 14-3-3 and not covalently cross-linked dimeric 14-3-3 (Fig. 3), indicating that 14-3-3 phosphorylated on Ser 58 is monomeric. The absence of cross-linked phosphorylated 14-3-3 dimers re-affirms the previous result ( Fig. 2A), showing that phosphorylation of a single monomer within a dimer is sufficient to disrupt the dimeric structure of 14-3-3. Therefore, Ser 58 phosphorylation has a dominant effect with respect to dimer disruption, increasing the potency of the event.
14-3-3 Phosphorylated on Ser 58 Retains the Ability to Bind Phosphoserine Peptide-Serine 58 of 14-3-3 is in the middle of helix 3 and from the crystal structure is buried at the dimer interface (7). Other residues in helix 3 are involved in client protein binding in the amphipathic groove; specifically Arg 56 and Arg 60 that flank Ser 58 have been demonstrated to be critical for binding to phosphoserine peptide (14). Therefore, it is important to determine whether phosphorylation of Ser 58 disrupts the amphipathic groove and consequently the ability of 14-3-3 to bind to phosphoserine client proteins. Thus, the ability of 14-3-3 phosphorylated on Ser 58 to bind a peptide corresponding to a 14-3-3 binding motif in the cytoplasmic domain of the granulocyte-macrophage colony-stimulating factor receptor ␤ chain (12) was determined. Phosphorylated and unphosphorylated peptides (16) were used to pull-down 14-3-3 from SDK reactions carried out either in the presence or absence of DMS (Fig. 4). Immunoblotting of the pull-down material with anti-14-3-3 antibody showed that 14-3-3 bound to the phosphorylated but not the unphosphorylated peptide, demonstrating that interaction of 14-3-3 with peptide is dependent on phosphorylation of the peptide. Similarly, Ser 58 -phosphorylated 14-3-3 also bound to phosphorylated peptide and, as expected, was detected only in SDK reactions carried out in the presence of DMS (Fig. 4). This indicates that Ser 58 -phosphorylated monomeric 14-3-3 retains the ability to bind to a 14-3-3 binding site and that binding is dependent on phosphorylation of the site, thus demonstrating that the amphipathic binding groove in the phosphorylated 14-3-3 monomer is not structurally perturbed by phosphorylation. FIG. 2. 14-3-3 phosphorylated on Ser 58 exhibits faster migration on native-PAGE relative to unphosphorylated 14-3-3. S100 extract from PMA stimulated Balb/c 3T3 fibroblasts was incubated either without (Ϫ) or with wild-type (ϩ) recombinant 14-3-3 (r14-3-3) in the presence or absence of DMS for 15 min at 37°C and proteins separated on 12.5% native-PAGE. After Western transfer the gel was immunoblotted with either anti-phospho-Ser 58 (A) or anti-14-3-3 antibody (B). C shows the mobility of purified recombinant 14-3-3 run on native-PAGE and stained with Coomassie Blue. The mobility of prestained protein markers is shown between the panels to compare migration of detected species.

14-3-3 Phosphorylated on Ser 58 Is Monomeric Phosphorylation and Monomerization of 14-3-3 Can Be Induced in Vivo by
Sphingolipids-Having demonstrated that phosphorylation of Ser 58 by SDK renders 14-3-3 monomeric in vitro, it is important to show that this process also occurs in vivo. We have examined the phosphorylation and dimerization of 14-3-3 in vivo in Balb/c 3T3 cells after sphingolipid treatment using the phospho-specific Ser 58 antibody (Fig. 5A). DMS was employed for these studies being a naturally occurring nonmetabolizable analogue of sphingosine. Whole cell lysates from Balb/c 3T3 cells treated with 25 M DMS for varying lengths of time were run on native-PAGE and immunoblotted with phospho-specific Ser 58 and anti-14-3-3 antibodies (Fig. 5A). The phospho-specific Ser 58 antibody detected a protein migrating with the seventh protein marker after 10 min of treatment with DMS corresponding to monomeric 14-3-3. This demonstrates that DMS treatment of Balb/c 3T3 fibroblasts does indeed disrupt dimeric 14-3-3 in vivo presumably by inducing phosphorylation of Ser 58 .
We extended these studies to other cell types and investigated the effect of DMS treatment on 14-3-3 phosphorylation and dimeric status in the murine lymphoid cell line, CTLL-2. Cells were stimulated with increasing doses of DMS for 30 min and whole cell lysates prepared and run on native-PAGE. After Western blotting the lysates were immunoblotted with either the phospho-specific Ser 58 antibody or anti-14-3-3 antibody (Fig. 5B). As can be seen, when cells were treated with 25 M DMS, phosphorylated 14-3-3 was detected with the Ser 58 phospho-specific antibody migrating exclusively with the seventh protein marker (Fig. 5B). This indicates that sphingolipid-induced phosphorylation of 14-3-3 on Ser 58 disrupts the dimeric structure of the endogenous protein in CTLL-2 cells. Furthermore, as seen in vitro, no phosphorylated 14-3-3 is detected as a dimer, indicating that phosphorylation of a single monomer within a 14-3-3 dimer is sufficient to disrupt the dimer. DISCUSSION We present data here that the dimeric status of 14-3-3 protein is regulated by site-specific phosphorylation. A sphingolipid-activable cytosolic kinase activity, previously described as SDK, phosphorylates 14-3-3 in vitro on a serine residue (Ser 58 ) that lies in the dimer interface of the protein (10). We have generated a phospho-specific antibody that recognizes 14-3-3 phosphorylated on Ser 58 , and this unique tool has allowed us to probe the effect of SDK-mediated phosphorylation of this site on 14-3-3 dimeric structure. We have employed native-PAGE and cross-linking studies to qualitatively discriminate between monomeric and dimeric forms of 14-3-3 and have found that sphingolipid induced phosphorylation of Ser 58 disrupts the dimeric structure of 14-3-3 both in vitro and in vivo (Figs. 2, 3,  and 5). Furthermore, phosphorylation of a single monomer unit within a dimer of 14-3-3 is sufficient to disrupt the dimer, increasing the potency of the event. We show that phosphorylated monomeric 14-3-3 is competent at phosphoserine peptide binding, thereby demonstrating that phosphorylation does not have gross conformational effects on the amphipathic phosphoserine binding groove. We propose that this regulated phosphorylation and monomerization of 14-3-3 controls it's deforming activity and may have profound implications for 14-3-3 function and the regulation of many cellular processes.
Phosphorylation of Ser 58 by SDK is not restricted to the 14-3-3 isoform. The original SDK report also demonstrated phosphorylation of the equivalent residues in ␤ and (10), and equivalent sites are found in all isoforms with the exception of and . Moreover, the amino acid sequence surrounding the serine in all phosphorylatable isoforms is highly conserved, suggesting that the ability to regulate monomerization by phosphorylation has been functionally conserved. Interestingly, not only is the sequence in the corresponding region of 14-3-3 and 14-3-3 more divergent but the homologous residue to Ser 58 is an alanine and therefore not susceptible to phosphorylation. It is intriguing that these more tissue-restricted isoforms are not phosphorylatable at this site and suggests that they may perform specific roles for which regulation of monomerization is not desirable.
Our findings are at odds with a recent report in which phosphorylation of Ser 58 in 14-3-3 by PKB/Akt was said to have no effect on dimer stability (13). However, as noted by others (8), the recombinant 14-3-3 used as kinase substrate in those studies already comprised a significant amount of monomeric 14-3-3, as determined by native-PAGE analysis. The extent of phosphorylation obtained coincided with the amount of monomer present, suggesting that 14-3-3 monomer rather than dimer may have been the substrate for Akt phosphorylation (13). In contrast, the recombinant 14-3-3 used in our in vitro studies is solely dimeric as determined by native-PAGE anal- In both cases whole cell lysates were then prepared and run on native-PAGE prior to Western transfer and immunoblotting (IB) with anti-phospho-Ser 58 (anti-pS58) or anti-14-3-3 antibody. The mobility of prestained protein markers is shown between the panels to compare migration of detected species. ysis, with no monomeric component detectable by immunoblotting (Fig. 2B) or Coomassie staining (Fig. 2C). Additionally, we detect monomeric 14-3-3 in vivo using the phospho-specific Ser 58 antibody (Fig. 5), indicating that the monomerization observed is physiological.
It is not possible from our studies to gauge the exact stoichiometry of 14-3-3 phosphorylation in vivo as the native-PAGE system used to discriminate monomers from dimers resolves only proteins and complexes with relatively acidic isoelectric characteristics. As shown (Fig. 4), phosphorylated 14-3-3 is still able to bind to phosphoserine peptide and so would therefore be expected to interact with phosphoserine client proteins and the resulting complexes may not resolve on native-PAGE. Therefore, the amount of phosphorylated 14-3-3 detected under these conditions may underestimate the level of phosphorylation actually occurring.
The proposed functional role for Ser 58 in regulating monomerization is consistent with its location in the dimer interface (7). However, as previously noted (8), how a kinase gains access to this buried residue remains unknown. The dimeric form of the protein is inherently stable and not prone to dissociation (9), and we have found no evidence for free exchange of monomers in mixing experiments using differentially tagged forms of 14-3-3 (data not shown). An alternative explanation could be that 14-3-3 protein is flexible, allowing access of SDK activity to the dimer interface, but evidence suggests that this is not the case as the crystal structures of 14-3-3 unbound or bound to a protein are remarkably similar (17). It is conceivable that sphingolipid may have some direct effect on 14-3-3 that allows access of the kinase to Ser 58 . Such modulation of Src kinase substrates by sphingolipids has been observed previously (18). Alternatively, in the original description of SDK activity, it was shown that the enzyme activity co-purified with 14-3-3 and could be co-immunoprecipitated with endogenous 14-3-3 (10), indicating that the kinase responsible for SDK activity is in intimate contact with 14-3-3, even in the absence of sphingolipid. Therefore, SDK may gain access to Ser 58 by conformationally altering 14-3-3 directly.
Previous studies with 14-3-3 mutants defective in dimerization indicate that although monomeric forms of 14-3-3 are capable of binding to phosphoserine target proteins, in many cases the monomer is unable to regulate client protein functions (19 -22). Furthermore, 14-3-3 mutants that are defective in client protein binding owing to substitutions in the phosphoserine binding amphipathic groove exhibit dominant negative effects on several pathways due to their ability to dimerize with endogenous 14-3-3, resulting in functionally monomeric forms (23,24). In the light of these studies we believe that the regulated monomerization of 14-3-3 demonstrated here has significant implications for the regulation of many 14-3-3 functions in cells.
To date, the most clearly defined role of 14-3-3 is to protect cells against apoptosis. A peptide with high affinity for 14-3-3, difopein, induces apoptosis when expressed in cells, due to its ability to compete with cellular proteins for 14-3-3 binding (25). Strikingly, expression of a functionally monomeric dominant negative form of 14-3-3 in mouse fibroblasts enhances the cells apoptotic response to UVC irradiation, serum withdrawal, and tumor necrosis factor-␣, highlighting the role of dimeric 14-3-3 in protecting cells against apoptosis (23). Therefore, regulated monomerization may play a role in inducing apoptosis by disrupting the anti-apoptotic function of dimeric 14-3-3. Consistent with this hypothesis, the sphingolipids sphingosine and DMS have both been shown to induce apoptosis in many cell types, including CTLL-2 cells (15, 26 -31). Indeed, under the conditions used here to induce 14-3-3 phosphorylation in CTLL-2 cells (Fig. 5B), significant cell death was observed after 5 h of treatment with DMS (data not shown). Therefore, it is conceivable that phosphorylation and consequent monomerization of 14-3-3 contributes to apoptosis induced by these sphingolipids and may have a more general role in functions where 14-3-3 is involved.